Using the data is made harder by not providing the entity number for linked records

[colmjude]

On the platform we link between related entities.

We must generate this link by working out or keeping a record of the related entity.

For example, this tree entity (7002019928) links to the parent TPO

The href for this link is https://www.planning.data.gov.uk/entity/7001003254
So the platform knows this tree (7002019928) links this TPO (7001003254).

Thats great.

But if I want to use the data I get the following JSON

{
    "reference": "15/TPO/011/T1",
    "prefix": "tree",
    "name": "Street Tree Outside Whitings Hill Primary School, Whitings Road, Barnet, EN5 2QY Tree Preservation Order 2015",
    "dataset": "tree",
    "organisation-entity": "48",
    "start-date": "2015-07-06",
    "end-date": "",
    "entry-date": "2025-02-01",
    "typology": "geography",
    "geometry": "MULTIPOLYGON (((-0.217726 51.645539, -0.217715 51.645544, -0.217716 51.645553, -0.217741 51.645552, -0.21774 51.645543, -0.217726 51.645539)))",
    "point": "POINT (-0.217728 51.645547)",
    "entity": 7002019928,
    "tree-preservation-order": "15/TPO/011",
    "tree-preservation-order-tree": "Individual"
}

How does a user of the data know what TPO entity it links to?

We should include the linked entities in the data payload.

Feels like consumers of the data are not currently benefiting from a usp of the platform; that it links datasets together.

[colmjude]

I'm sure I'm missing something obvious but would it not be as simple as updating the JSON for the entity to include something link tree-preservation-order-entity. For example

{
    "reference": "15/TPO/011/T1",
    "prefix": "tree",
    "name": "Street Tree Outside Whitings Hill Primary School, Whitings Road, Barnet, EN5 2QY Tree Preservation Order 2015",
    "dataset": "tree",
    "organisation-entity": "48",
    "start-date": "2015-07-06",
    "end-date": "",
    "entry-date": "2025-02-01",
    "typology": "geography",
    "geometry": "MULTIPOLYGON (((-0.217726 51.645539, -0.217715 51.645544, -0.217716 51.645553, -0.217741 51.645552, -0.21774 51.645543, -0.217726 51.645539)))",
    "point": "POINT (-0.217728 51.645547)",
    "entity": 7002019928,
    "tree-preservation-order": "15/TPO/011",
    "tree-preservation-order-entity": 7001003254,
    "tree-preservation-order-tree": "Individual"
}

Or would something like this be better?

{
    ...
    "geometry": "MULTIPOLYGON (((-0.217726 51.645539, -0.217715 51.645544, -0.217716 51.645553, -0.217741 51.645552, -0.21774 51.645543, -0.217726 51.645539)))",
    "point": "POINT (-0.217728 51.645547)",
    "entity": 7002019928,
    "tree-preservation-order": {
        "reference": "15/TPO/011",
        "entity": 7001003254
    },
    "tree-preservation-order-tree": "Individual"
}

[RichLimbrick]

If I'm understanding correctly, where the entity is the true unique ID for the record, then that's what we need to be included to link the records across datasets. So your example of tree-preservation-order-entity being included is what makes sense to me. As a more non-technical user of the data, I would be able to work out what was going from that.